126 JOURNAL OF ECONOMIC ENTOMOLOGY [Vol. 1 



be urged that natural selection will develop a long ovipositor in any 

 group where it may be of service, but the value of the elongated ovi- 

 positor for the separation of larger groups shows that it is an organ 

 that is very slightly influenced by specific exigencies. 



Seasonal distribution naturally limits the range of parasitism rather 

 closely, since the time of appearance of insect species varies greatly; 

 and as adult parasites are not long-lived, their hosts must appear at 

 nearly the same time as they do themselves. The tendency is for 

 parasitic species to undergo metamorphosis more rapidly than others, 

 probably on account of their easily assimilated food, and with shorter 

 life-cycle they will tend to pass through more generations in a season 

 than their hosts in many instances. Here I believe lies the reason for 

 the acquirement of more than a single host by some species, although 

 it cannot explain the large number of hosts of some species, nor the 

 several closely related hosts of others. 



While some of the probable reasons for the association of host and 

 parasite may be found among the foregoing, there seem to be no cases 

 that can be fully elucidated by either one or a combination of all. 

 Indeed, the very fact that there are so many closely related parasites 

 and so little transfer to new hosts, would almost preclude such a sup- 

 position a priori, and some more subtle hereditary influence must be 

 sought for. 



Throughout the entire parasitic group, there are scattered here and 

 there species which are particularly adaptive, in that they attack sev- 

 eral or a number of more or less closely related hosts, while others not 

 far removed taxonomically have apparently but a single one. But in 

 nearly every minor group and in some of the larges ones, there is a 

 well-marked tendency to select as hosts species belonging to another 

 homogeneous group. Thus every one of the seven or eight hundred 

 members of the Proctotrypid family Scelionidce, so far as known, with 

 one single exception of doubtful relationship, is an egg-parasite. Al- 

 most every species of the family Alysiidce is parasitic on dipterous 

 larvae ; practically all members of the extensive sub-family Aphidiince 

 of the Braconidce are aphid parasites, and so we might continue to 

 list many more. There are here also very noticeable exceptions, but 

 they only serve to show the strong tendency toward uniformity which 

 exists everywhere. 



Small groups do not always show the increased specialization which 

 we might expect from the uniformity exhibited by so many larger 

 groups, and species in particular may occasionally have almost as 

 great a range in choice of hosts as genera or even larger groups. A 

 case in point is the Chalcidid genus TricJiogramma, which attacks the 



I 



